Composite Structures 152 (2016) 737–745

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Development and qualification of a new polymeric matrix laminated

composite for pipe repair
N.R.F. Rohem a, L.J. Pacheco b, S. Budhe b, M.D. Banea b, E.M. Sampaio c, S. de Barros b,⇑
a
Instituto Federal Fluminense, Rodovia Amaral Peixoto, s/n-Macaé, 27932-050 Rio de Janeiro, Brazil
b
Federal Center of Technological Education Celso Suckow da Fonseca-CEFET/RJ, Av. Maracanã, 229, 20271-110 Rio de Janeiro, RJ, Brazil
c
Universidade do Estado do Rio de Janeiro-UERJ, Rua Bonfim, 25, Nova Friburgo, 28625-570 Rio de Janeiro, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Nowadays the use of polymeric matrix composites to repair and strengthen the damaged pipe structures
Received 24 May 2016 in the oil industry has become a common practice. Hence, it is essential to validate the performance of
Accepted 26 May 2016 new developed composite laminate materials for the damaged pipe repair. In this study, the effectiveness
Available online 28 May 2016
of a new composite laminate for the pipes repair was investigated. First, the mechanical and thermal
properties of the new developed composite laminate were determined. Next, two defect types, Type A
Keywords: (non-through wall) and Type B (through wall) were manufactured into the pristine pipe specimen and
Pipe repair
the evaluation of the performance of the repaired pipe was carried out by hydrostatic tests. The perfor-
Hydrostatic test
Polymeric matrix composites
mance of the repaired pipe using the new developed composite laminate material was satisfactory in
Metal-composite repair both defect types.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction the hotwork [5,6]. Recently, fiber reinforced polymer (FRP) matrix
composite repair systems have emerged as an alternative repair
Pipe and pipelines are extensively used to carry/transport fluid system to the welding technique. The application of adhesive
(oil, gas, water) over a long distance. Generally, the pipe is made up bonded joints in structural components made of fiber-reinforced
of steel metal due to its high strength, low cost, efficient and safe composites have increased significantly in the last years [7].
compared to other materials [1]. However, it is more sensitive to Repairs made with FRP materials offer distinct advantages such
corrosion in harsh environment, particularly in the presence of as reduced cost [8], corrosion prevention [8–10], more safety [11]
sea water and sulfur ingress media [2,3]. As is known, the pipelines and quick repair [10] etc.
are exposed to a harsh environment such as submerged in water The use of fiber reinforced composites has already been proven
(offshore unit), underground pipe (sewage pipelines, oil pipelines) an effective tool for the repair of damaged structures [12–15]. To
and above the ground pipe (water pipelines). Almost, in all practi- evaluate the effectiveness of a given repair system, it is important
cal applications the pipes are subjected to harsh environmental to study the critical parameters that affect the performance of the
conditions which lead to an internal and an external damage, spe- repair composite (i.e. geometry of the repair including thickness,
cially corrosion damage and deterioration. An external damage type of composite materials including fiber and resin, orientation
also occurs in the form of cracking, wear, dents caused by impact of fibers and method of installation etc). [13,15,16]. In the last
during transportation and installation process [4]. In the past, it years, several researchers [17–21] investigated the mechanical
was common practice to completely replace the affected sections and thermal properties of the composite repair material in order
of pipe. However, as this always involves a transport stoppage, to have the best possible combination of composite material and
repair systems are seeing as fast and economical alternatives as assure the repair performance. Successful application of composite
do not interrupt operations. repair material for the corroded or damaged pipe was found
Traditionally, welding technique is used to repair the damaged [22–25]. However, the pipeline repair still presents some difficul-
part by cutting or replacing the damaged section and use of a steel ties regarding its long term durability. These difficulties are related
patch. However, this technique requires to stop the operation to the lack of a sufficient number of test results and the validation
while the repair is being performed and the process also involves of the composite material properties.
The present paper deals with an experimental analysis of a new
⇑ Corresponding author. glass fiber reinforced repair system for metallic pipelines with a
E-mail address: silvio.debarros@gmail.com (S. de Barros). standard defect size as per the standard ISO/TS24817 [26].

http://dx.doi.org/10.1016/j.compstruct.2016.05.091
0263-8223/Ó 2016 Elsevier Ltd. All rights reserved.
738 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745

The primary focus of this research program was the development

of the glass-based composite repair system and further the system
was tested and validated by the standard ISO/TS24817 [26]. Hydro-
static tests were performed to validate the performance of the
composite repair system.

2. Materials and methods

2.1. Materials

2.1.1. Woven fabric

The selection of woven fabric composite material for tube repair
should have the following characteristics: light-weight which offer
in greater ease for resin impregnation. The resin-woven proportion
used was 2:1. The fabric should have approximately 66% of their
fibers orientated in its longitudinal direction (circumferential
direction of the duct) and 34% in the transverse direction (axial
duct direction). The choice of this setting is based on the stresses
acting on a thin-walled cylinder (circumferential stress is equal
to 2 times of axial stress). A bidirectional fabric of glass fibers ori-
ented at 0° in its longitudinal direction and 90° to the transverse
direction was used. A weight of 434 g/m2 and 261 g/m2 to the lon-
gitudinal and transverse direction, in percentage 62% and 38%,
respectively was used. Fig. 1. Flat plate preparation process for the specimens.

2.1.2. Laminating resin composite laminate was characterized using DMA (Q800, TA
For the selection of laminating resin the following factors were instrument). The range of testing temperature was controlled from
taken into account: fiber wettability, viscosity, curing time and 40 °C to 250 °C with a heating rate of 2 °C/min.
glass transition temperature (Tg). A bi-component epoxy resin,
PIPEFIXÒ developed by Novatec ltd (spin-off of Laboratório de 2.2.3. Tensile test
Adesão e Aderência – UERJ, Nova Friburgo RJ, Brazil). The curing The flat plate was kept in an oven for the post curing process at
time was 2 h, followed by a 1 h post cure at 130 °C. A heating ramp 130 °C for 1 h. After post curing, the flat plate was cut at required
of 3 °C per minute was used. The properties of the constituent dimensions for tensile test specimen as per the standard ASTM
materials used in the hand lay-up process supplied by the manu- D3039 [29]. Tensile specimens were conditioned at 23 °C and
facturer are presented in Table 1. 50% relative humidity for a minimum of 24 h.
The specimens were tested in a universal testing machine (Shi-
2.1.3. Steel pipe madzu AGI 100 kN) at room temperature and relative humidity of
The pipe material was an API-5L X56 steel with the following 50% ± 10%. The test speed was 2 mm/min. Five specimens were
basic properties: Younǵs Modulus Epipe = 210 GPa; yield stress tested at each condition. The load–deformation curve of specimens
Sy = 450 MPa and ultimate strength = 627 MPa. obtained from the machine and an extensometer (model SG50-50
Shimadzu) was recorded. Fig. 4 shows the tensile test setup with
2.2. Methods an extensometer attached to the specimen.

2.2.1. Acid digestion test 2.2.4. Hydrostatic test

A flat plate was manufactured by hand lay-up technique and it The specimens for the hydrostatic test were fabricated accord-
consists of four stacking layers of laminate. The flat plate prepara- ing to the procedure described in ISO/TS24817 standard [26]. In
tion process is shown in Fig. 1. A sample of approximately 0.3 g this work, two types of defect were studied, one is defect Type A,
weight was cut from the plate before the post curing process for in which the substrate is not leaking and not expected to leak
the acid digestion test in order to determine the percentage of fiber within the lifetime of the repair system, requiring structural rein-
and resin by weight. forcement only and defect Type B, in which the substrate requires
The acid digestion test was performed according to the proce- structural reinforcement and sealing of through-wall defects
dure described in ASTM D3171 [27]. Fig. 2 shows the stepwise pro- (leaks).
cedure for acid digestion test.
2.2.4.1. Defect Type A. The test samples were prepared from carbon
2.2.2. Thermal test (glass transition temperature test) steel API 5L X56 SCH40 pipe sections with a nominal diameter of
Eight samples were cut out from the same flat plate for the 152.4 mm and axial length of 600 mm. The carbon steel pipe,
determination of glass transition temperature. The dimension of SCH40 designation corresponds to a wall thickness (t) of
each sample was 20
12.5
1.5 mm in accordance with ASTM 7.11 mm and an outer diameter (D) of 168.3 mm. The tube speci-
D7028 standard [28] (see Fig. 3). Glass transition temperature of mens were prepared by machining the defect into the pipe to sim-
ulate an 80% external wall loss thickness. The defect was machined
by electrical discharge machining (EDM). The basic dimensions of
Table 1
Material properties used for manual lamination. pipes and defect size are shown in Fig. 5a.
The defect area of steel tube (Fig 5b) was initially degreased
Material Density (g/cm3) Youngs modulus (GPa)
with water and then cleaned with sandpaper (P100), in order
Fiber glass 2.55 72 to remove oil residues coming from the machining process and
Epoxy resin 1.18 3.5
promote a better adhesion between pipe and composite. Initially
 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745 739

Fig. 2. Acid digestion test setup and step up wise procedures [(a) sample weight, (b) sample placed in a heated bath of sulphuric acid, (c) H2O2 added to hot mixture, (d) after
sufficient time the mixture becomes transparent (e) solution containing the fibers and (f) filtration of the fibers with the aid of a vacuum pump].

the adhesive was applied to the defect area and then a thin layer of
epoxy was applied over the entire circumference. The calculation
of composite laminate repair thickness in accordance with ISO/TS
24817 standard [26], was done using the following equation:
 
1 Pf D
trepair ¼  sa t s ð1Þ
Ec eshort 2
where trepair = repair thickness (mm), Ec = Circumferential modulus
of the laminate repair (MPa), eshort = deformation limit on short-
term laminate repair, Pf = failure pressure (MPa), D = tube diameter
(mm), sa = tube yield strength (MPa), ts = remaining wall thickness
of defect after machining (mm).
54 turns of laminates were needed to complete the 16 mm
repair thickness calculated as per standard and each turn was fol-
lowed by application of the resin. Fig. 6 shows the repair specimen
preparation for defect Type A.
Hydrostatic pressure tests were performed according to the
procedure described in the technical report of pipeline specimen
repair [30] and ISO/TS 24817 standard [26].
The hydrostatic tests were carried out in three steps:

Fig. 3. Flat plate cut into the different size of specimen for DMA analysis. 1st Step: first cycle of 4 h at a pressure of 1.25
Pmax.

Pmax is the maximum operating pressure and is defined by the

following equation:
t
Pmax ¼ 0:72  S  ð2Þ
r
where S = pipe yield strength (MPa), t = nominal thickness of pipe
wall (mm), r = pipe radius (mm).

2nd step:10 dynamic cycles with a pressure range 0 to Pmax.

3rd step: The third step was performed at failure pressure (Pf ).
The failure pressure is calculated by the following equation:

2tsa
Pf ¼ ð3Þ
D
where Pf = failure pressure (MPa), t = nominal thickness of pipe wall
(mm), sa = pipe yield strength (MPa), D = external pipe diameter
(mm).

2.2.4.2. Defect Type B. A laminate repair was designed to be capable

Fig. 4. Tensile test set up with an extensometer. of restoring the structural integrity of the pipe having a through wall
740 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745

(a) Design (all dimensions in mm).

(b) Photo of the steel tube with defect prepared by EDM machining
Fig. 5. Pipe and defect geometry.

Fig. 6. Specimen repair preparation of defect Type A for hydrostatic testing.

 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745 741

Table 2 Law of mixture E1 ¼ Ef V f þ Em V m ð4Þ

Volumetric percentage of fiber and resin.

Determination of volumetric percentage of fiber 1 Vf Vm

Shear modulus ¼ þ ! G12 ¼ 1630:0 MPa ð5Þ
Initial sample weight (g) 0.3154 G12 Gf Gm
Weight after digestion (g) Fiber Resin
0.1259 0.1895 Poisson ration t12 ¼ tf V f þ mm V m ! m12 ¼ 0:28 ð6Þ
Percentage by weight (%) Fiber Resin
39.92 60.08 In case of a two-dimensional configuration, the fibers oriented in the
Percentage (%) Fiber Resin longitudinal direction are considered as a booster, and the fibers in
23.51 76.49
the transverse direction are considered as the matrix. Based on the
weight of the fabric, the laminate has 62.45% of the fibers in the cir-
cumferential direction and 37.55% of the fibers in the axial direction.
Table 3 From Table 3, it’s clearly seen that the difference between the
Theoretical and empirical value of Young’s modulus in both directions (circumfer-
theoretical and empirical results is minimum, which indicate the
ential and axial).
efficient manufacturing process of the specimens.
Fiber volumetric Young’s Young’s Different
percentage (%) modulus modulus in modulus
3.2. Glass transition temperature (Tg)
theoretical (GPa) empirical (GPa) (%)
Circumferential 14.68 13.56 13.13 3.17
The glass transition temperature was determined by three
Axial 8.83 9.55 8.86 7.23
visco-elastic parameters:the storage modulus, the loss modulus,
and the tan delta method. Fig. 7 shows the glass transition temper-
defect. To represent the typical Type B defect a circular hole was ature value by all methods. The summary of the measured glass
drilled through the wall thickness of pipe. Section of metallic pipe transition temperatures is shown in Table 4. The storage modulus
of minimum diameter 100 mm and minimum thickness of 3 mm glass transition temperature (Tg = 116 °C) was selected for the
was used as per the annexure D of standard ISO/TS 24817 [26]. 9 preparation of specimens based on the mechanical and hydrostatic
specimens were prepared with different hole size of 10 mm, testing trials.
15 mm and 20 mm. 3 specimens were manufactured of each hole
size. Hydrostatic tests were conducted at temperature, Tg-30 °C as 3.3. Tensile test results
per ISO/TS24817 standard [26]. The tubes were closed with a set
of flanges connected by rods. The ramp pressure increases approxi- The tensile data (young’s modulus and tensile strength values of
mately 0.1 MPa (1 bar) per second as per defined in ASTM D1599 glass fiber reinforced laminate in both directions circumferential
[31] standard. Pressure increases until a delamination occurs and axial respectively), were calculated from the stress–strain
between the composite laminate layers and the substrate. curves recorded and the results can be seen in Tables 5 and 6,
respectively.
3. Results and discussions
3.4. Hydrostatic test results
3.1. Volumetric percentages of fiber and resin
3.4.1. Defect Type A
The results obtained from the acid digestion tests are presented The pressure and temperature curves of the hydrostatic test for
in Table 2. The acid digestion test provides the percentage of defect Type A are presented in Figs. 8–11.
weight of fibers and resins.
Shear modulus and poisson’s ratio are calculated from the law 1st step: first cycles of 4 h at a pressure of 1.25
Pmax, i.e.
of mixture and data from the acid digestion test. 293.6 bar.

Fig. 7. Glass transition temperature value by three methods (storage modulus, loss modulus, and tan delta).
742 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745

Table 4 2nd Step: 10 cycles with a pressure range between 0 and

Measured glass transition temperatures determined by the storage modulus, loss 234.9 bar
modulus peak, and tan delta peak methods.
3rd step: the third step at failure pressure (Pf) 326.2 bar.
Storage modulus [°C] Loss modulus peak [°C] Tan delta peak [°C]
116.02 122.77 134.76 Fig. 8 represents the pressure and the temperature curve of all
three steps of hydrostatic test. Pressure and temperature curve of
the individual step of hydrostatic test is presented in Figs. 9–11.
Table 5 Test results showed that the repair systems sustained the orig-
Tensile test results of glass fiber reinforced laminate along circumferential direction. inal design pressure. This system operated during a minimum of
Tensile test direction 1 (circumferential)
four hours under a hydrostatic pressure test and sustained 1.389
times above the design pressure and was able to withstand ten
CP Thickness Width Gauge length Max. tension Young’s
(mm) (mm) (mm) (MPa) modulus (GPa)
pressure cycles of zero-to-design pressure without showing any
visual damage (Fig. 14).
1 4.6 24.8 50 219.33 12.09
2 4.6 23.7 50 251.44 13.58
Da Costa Mattos et al. [22] also observed that the calculated
3 4.7 22.5 50 270.81 13.25 theoretical failure pressure was lower than the internal pressure
4 4.6 23.6 50 260.19 13.83 as recommended by ISO 24817 [28] for the repaired system that
5 4.7 23.6 50 260.19 12.90 they investigated. However, the authors [22,23] suggested that
Average 252.39 13.13 more experimental tests are needed in order to confirm and vali-
Standard deviation 19.72 0.68 date the results.

3.4.2. Defect Type B

Table 6 Hydrostatic tests were conducted with heated water at Tg-30 °C
Tensile test results of glass fiber reinforced laminate along axial direction.
(116–30), corresponding to a 86 °C test temperature as defined in
Tensile test direction 2 (axial) the standard. Initially, the water was heated up to 60 °C with the
CP Thickness Width Gauge length Max. tension Young’s valve open to maintain the atmospheric pressure. A 10 bar pres-
(mm) (mm) (mm) (MPa) modulus (GPa) sure was maintained to avoid vapor formation after reaching
1 5 23.1 50 164.51 8.80 approximately to 60 °C. Fig. 12 shows the pressure and tempera-
2 5 22.6 50 153.61 9.23 ture curve during the hydrostatic test of the specimen with 10,
3 5 23.6 50 161.94 8.83 15 and 25 mm hole size specimens. A higher pressure was sus-
4 5 21.6 50 155.89 8.98
tained by the lower hole size (10 mm) defect specimen and then
5 4.9 23 50 159.21 8.46
15 mm and last 25 mm as expected before the delamination
Average 159.03 8.86
failure.
Standard deviation 4.41 0.28

Fig. 8. Pressure and temperature curve of hydrostatic test of defect Type A.

Fig. 9. Pressure and temperature curve of hydrostatic test of defect Type A (step1_4 h at 1.25*Pmax).
 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745 743

Fig. 10. Pressure and temperature curve of hydrostatic test of defect Type A (step2_10 pressure cycles within limit 0 to design pressure).

Fig. 11. Pressure and temperature curve of hydrostatic test of defect Type A (step3_
failure pressure).

Fig. 13. Measured failure pressure for hydrostatic testing of through wall defect
configuration.

Fig. 12. Pressure and temperature curve of 10 mm, 15 mm and 25 m hole size
defect specimen during hydrostatic test of defect Type B.

The results of through wall defects specimen with different hole Fig. 14. The tube specimen with plastic deformation at both ends.
size defects are shown in Fig. 13. Each point represents an average
value of failure pressure and the error bars the standard deviation
for each batch of three specimens. As expected, smaller hole diam- 4. Failure analysis
eter defect lead to sustain higher failure pressure due to the
reduced blister deformation at a given pressure. The same range 4.1. Defect type A
of failure pressure was observed by Alnaser et al. [32], for the same
hole size diameter and repair thickness of specimen repaired using Fig. 14 shows the tube specimen after applying the failure pres-
carbon fiber composite materials. Mableson et al. [33] also found a sure. Plastic deformation was observed at both ends of the pipe
similar failure pressure range even though the repair thickness without failure in the composite repair section. On magnifying
increased twice. the image, it is clearly seen that the plastic deformation of the tube
744 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745

Fig. 15. Failure specimen, defect Type B.

occurred instead of failure of the composite repair. Hence, the com- [6] Koch GH, Brongers MPH, Thompson NG, Virmani YP, Payer JH. Corrosion cost
and preventive strategies in the United States. Federal Highway Administration
posite repair with the repair thickness (16 mm) can sustain the
Office of Infrastructure Research and Development; 2001. p. 260–311.
desired pressure without failure of the composite repair pipe. [7] Banea MD, da Silva LFM. Adhesively bonded joints in composite materials: an
overview. Proc IME J Mater Des Appl 2009;223:1–18.
4.2. Defect Type B [8] Seica VM, Packer AJ. FRP materials for the rehabilitation of tubular steel
structures for under water applications. Compos Struct 2007;80:440–50.
[9] Cuthill J. Advances in materials, methods, help gain new users. Pipeline Gas J
The failure occurs (Fig. 15a and b) at the interface between the 2002;229(11):64–6.
laminate and the substrate as per standard ISO/TS 24817 [26] (ISO [10] de Barros S, Banea MD, Budhe S, de Siqueira CER, Lobão BSP, Souza LFG.
Experimental analysis of metal-composite repair of floating offshore units
specify that the failure cannot occur along the thickness of the lam- (FPSO). J Adhesion 2016. http://dx.doi.org/10.1080/00218464.2016.1177514.
inate and must take place only by delamination). Red colorant was [11] Marsh G. Composites renovate deteriorating sewers. Reinf Plast 2004;48
added to the test water in order to highlight the path taken by the (6):20–4.
[12] Shamsuddoha M, Islam MM, Aravinthan T, Manalo A, Lau KT. Effectiveness of
water at the time of failure. From Fig. 15a it can be seen that the using fibre-reinforced polymer composites for underwater steel pipeline
failure occurred by delamination on the interface between the pri- repairs. Compos Struct 2013;100:40–54.
mer and the first layer of glass fiber and the leakage occurred at the [13] Shamsuddoha M, Islam MM, Aravinthan T, Manalo A, Lau KT. Effectiveness
characterisation of mechanical and thermal properties of epoxy grouts for
edges of the laminate. From Fig. 15b it can be seen that the leak composite repair of steel pipelines. Mater Des 2013;52:315–27.
occurred immediately below the first layer of the laminate repair. [14] Duell JM, Wilson JM, Kessler MR. Analysis of a carbon composite overwrap
pipeline repair system. Int J Pressure Vessel Pip 2008;85:782–8.
[15] da Costa Mattos HS, Paim LM, Reis JML. Analysis of burst tests and long-term
5. Conclusions hydrostatic tests in produced water pipelines. Eng Fail Anal 2012;22:128–40.
[16] Saeed N, Ronagh H, Virk A. Composite repair of pipelines, considering the
In this work, a new composite laminate material for pipe repair effect of live pressure-analytical and numerical models with respect to ISO/TS
24817 and ASME PCC-2. Compos Part B-Eng 2014;58:605–10.
was developed and the results were validated according to ISO/TS [17] Saint-Michel F, Chazeau L, Cavaillé JY, Chabert E. Mechanical properties of high
24817. The developed repair system is suitable to repair Type A density polyurethane foams: I. Effect of the density. Compos Sci Technol
(non-through wall) as well as Type B (through wall) defects. 2006;66:2700–8.
[18] Saint-Michel F, Chazeau L, Cavaillé JY. Mechanical properties of high density
The comparison of the composite laminate material properties polyurethane foams: II. Effect of the filler size. Compos Sci Technol
obtained through experimental and theoretical model showed a 2006;66:2709–18.
good correlation, which gives confidence of the procedures adopted [19] Husic S, Javni I, Petrovic ZS. Thermal and mechanical properties of glass
reinforced soy-based polyurethane composites. Compos Sci Technol
for the preparation of specimens. The performance of the repaired
2005;65:19–25.
tube using the new developed composite laminate material is satis- [20] Wilberforce S, Hashemi S. Effect of fibre concentration, strain rate and
factory in both types of defect under hydrostatic test as per ISO/ weldline on mechanical properties of injection-moulded short glass fibre
reinforced thermoplastic polyurethane. J Mater Sci 2009;44:1333–43.
TS24817 standard [26]. The results demonstrate that composite
[21] Bruckmeier S, Wellnitz J. Flexural creeping analysis of polyurethane
materials are a viable tool for repairing the damaged steel tube composites produced by an innovative pultrusion process. Sustainable
and can sustain the maximum possible pressure without failure. Autom Technol 2011;2:13–8.
[22] da Costa Mattos HS, Reis JML, Paim LM, da Silva ML, Lopes Junior R, Perrut VA.
Analysis of a glass fibre reinforced polyurethane composite repair system for
Acknowledgements corroded pipelines at elevated temperatures. Compos Struct
2014;114:117–23.
The authors would like to acknowledge the support of the [23] da Costa Mattos HS, Reis JML, Paim LM, da Silva ML, Lopes Junior R, Perrut VA.
Failure analysis of corroded pipelines reinforced with composite repair
Brazilian Research Agencies CNPQ, CAPES and FAPERJ. systems. Eng Fail Anal 2016;59:223–36.
[24] Djukic LP, Sum WS, Leong KH, Hillier WD, Eccleshall TW, Leong AYL.
References Development of a fibre reinforced polymer composite clamp for metallic
pipeline repairs. Mater Des 2015;70:68–80.
[25] da Costa Mattos HS, Reis JML, Sampaio RF, Perrut VA. An alternative
[1] Kennedy JL. Oil and gas pipeline fundamentals. 2nd ed. Tulsa,
methodology to repair localized corrosion damage in metallic pipelines with
Oklahoma: PennWell Publishing Company; 1993.
epoxy resins. Mater Des 2009;30:3581–91.
[2] Frankel GS. Pitting corrosion of metals a review of the critical factors. J
[26] ISO/TS 24817. Petroleum, petrochemical and natural gas industries. Composite
Electrochem Soc 1998;145:2186–98.
repairs for pipework. Qualification and design, installation, testing and
[3] Francis R. Galvanic corrosion of high alloy stainless steel in sea water. Br Corros
inspection; 2006.
J 1994;29:53–7.
[27] ASTM D 3171 Standard test methods for constituent of composite materials;
[4] Duell JM, Wilson JM, Kessler MR. Analysis of a carbon composite over wrap
2006.
pipeline repair system. Int J Pressure Vessel Pip 2008;85:782–8.
[28] ASTM D7028 Standard Test Method for Glass Transition Temperature (DMA
[5] Goertzen WK, Kessler MR. Dynamic mechanical analysis of carbon/epoxy
Tg) of Polymer Matrix Composites by Dynamic Mechanical Analysis (DMA);
composites for structural pipeline repair. Compos Part B-Eng 2007;38(1):1–9.
2009.
 N.R.F. Rohem et al. / Composite Structures 152 (2016) 737–745 745

[29] ASTM D3039 standard test method for tensile properties of polymer matrix [32] Alnaser IA, Keller MW. Comparison of coupon and full-scale determination of
composite material; 2008. energy release rate for bonded composite repairs of pressure equipment. Eng
[30] Freire JLF, Perrut VA, Braga AMB, Vieira RD, Ribeiro AS, Rosas MAP. Use of FBG Fract Mech 2015;146:31–40.
strain gages on a pipeline specimen repaired with a CFRE composite. Exp Tech [33] Mableson AR, Dunn KR, Dodds N, Gibson AG. Refurbishment of steel tubular
Soc Exp Mech 2015;39:70–9. pipes using composite materials. Plast Rubber Compos 2000;29(10):558–65.
[31] ASTM D1599: Standard test method for resistance to short-time hydraulic
pressure of plastic pipe, tubing, and fittings; 2005.